Stream Processing in the Big Data Era: a blast from the past

Time: 02:00 pm-03:00 pm
Location: Room Levico, Polo scientifico e tecnologico "Fabio Ferrari", Building Povo 2, via Sommarive 9, Povo (Trento)

Speaker  

Prof. Yannis Kotidis, Athens University of Economics and Business

Abstract

Big data applications are often characterized by the need to react quickly to information that is streamed by with high velocity. In this talk we will first discuss approximation techniques based on Locality Sensitive Hashing that have been successfully applied in sensor network monitoring applications. These techniques permit us to summarize in real time streamed sensor measurements and perform similarity computations among them. We will then show how the same techniques can be applied in the big data domain, enabling computation of similarity/correlation metrics in a distributed system. Time permitting, I will also discuss some recent results from our group related to online partitioning of large streaming graphs.

About the Speaker 

Prof. Yannis Kotidis is an Associate Professor in the Department of Informatics at the Athens University of Economics and Business. He holds a B.Sc. degree in Electrical Engineering and Computer Science from the National Technical University of Athens, a MSc and a Ph.D. in Computer Science from the University of Maryland (USA). Between 2000 and 2006 he was a Senior Technical Specialist at the Database Research Department of AT&T Labs-Research in Florham Park, New Jersey.  His main research areas include big data, data warehousing, data mining, graph databases and sensor networks. His most recent work focuses on management and analysis of big, streaming, complex datasets. He has published more than 80 articles in international conferences and journals and holds 5 U.S patents. He has served on numerous organising and program committees of international conferences related to data management. He has received the best paper award in ACM SIGMOD 1999 for his work in dynamic view management and the Research Excellence Award by AT&T Labs-Research (2001) for his work in large-scale data integration. He is also a co-author of the ICDE 2009 paper on streaming joins that received the Best Student Paper Award in that conference.

Contact person regarding this talk : Yannis Velegrakis, velgias [at] unitn.it